Yes, indeed it is a critical length if if is your desire to superimpose the bandpass curve properly on the pass curve of the duplexer. It should be an electrical 1/4 wave that accounts for the velocity propagation of the cable plus the electrical length of the coupling element in the Celwave cavity. If your end user doesn't have a tracking generator, IMO, attempting this is an exercise in futility. If all the bottles were built by the same OEM you could probably get a figure from their tech support group but with different OEMs you are
going to have to cut/add and try. First tune the duplexer for the desired pass and reject frequencies. Then tune the pass cavity for the desired pass frequency. Then glue it all together with an interconnect that guestimates a 1/4 wave including the coupling length in the pass cavity and look at on the tracking generator to see whether the pass curve gets steeper but remains essentially the same. It most likely will not. Add a couple of right angle adaptors to the interconnect and see if the pass curve distortion gets better or worse. If it's worse, shorten the interconnect cable and try again. If it gets better, lengthen the interconnect. Having said that, I think Skipp's point is well taken - if the junk is on channel, an additional pass cavity isn't going to eliminate it. BTW, are you using an isolator on the TX? K7IJ In a message dated 7/26/2007 5:11:16 A.M. Pacific Daylight Time, [EMAIL PROTECTED] writes: My question is whether the coax length is critical between the RX port of the Wacom duplexer and the input port of the Celwave cavity? I plan to send along a length of RG-393 (double shielded teflon coax) with the cavity. As far as I know, it is a random length. Should I cut it to something closer to 1/2 wavelength? 3/4 WL? ************************************** Get a sneak peek of the all-new AOL at http://discover.aol.com/memed/aolcom30tour
